Effects of Varying Ethanol and Turpentine Levels on Attraction of Two Pine Root Weevil Species, Hylobius pales and Pachylobius picivorus (Coleoptera: Curculionidae) L. K. RIESKE .\NO K. F. RAFFA Department of Entomology. University of Wisconsin. ~ladison. Wisconsin 53.06 EnvirOlL Entomol. 20< 1): 48-52 (1991) ABSTRAcr The pales weevil. Hylobius pales (Herbst). and the pitch-eating weevil. PachlJ­ lohim picicorus (Gennar), are part of a.weevil cOmplex causing extensive damage to plan­ tation pines throughout the Lake States. A means of monitoring w~l populations has been developed using pitfall traps baited with ethanol and turpentine. The relative attractiveness of six combinations of ethanol anc;I turpentine were compared. Traps were monitored through­ out the 1988 and 1989 growing seasons in a Scots pine Christmas tree farm. Both species were more strongly attracted to ethanol/turpentine ratios above 1:1. Pales weevils preferred slightly higher ethanol/turpentine ratios than did pitch-eating weevils. Within species. both sexes responded equivalently among treatments. The implications of these behavioral re­ sponses to weevils' locating stressed trees. the role of chemical ecology in niche partitioning, and IPM of pine root weevils are discussed. KEY WORDS Insecta. HlJlobius spp., PtichlJlobius spp., host volatiles THE PALES WEEVIL, Hylobius pales (Herbst), and Olfactory stimulation is an important aspect of the pitch-eating weevil Pachylobius picifJOTtJS host location and recognition (Mustaparta 1984). (Cermar), are part of a root-feeding complex that Volatile compounds emanating from host plant tis­ threatens plantation pine production in the Great sues are known to attract many insect species. Both Lakes region (Finnegan 1959, Mosher & Wilson ethanol and turpentine are recognized as important 1977). These species share a common host range host location and recognition cues in a variety of and are morphologically, behaviorally, and ec0­ pine-infesting insects (Moeck 1970, Fatzinger 1985), logically similar (Drooz 1985). Larvae of both spe­ including Hylobius and Pachylobius spp. (Tilles cies develop within roots of hosts weakened or re­ 1986a,b; Raffa &- Hunt 1988; Hunt & Raffa 1989). cently harvested. Adult weevils feed nocturna1ly . Turpentine is composed of host monoterpenes, of on pine branches and stems. This feeding can dam­ which a-pinene and p-pinene are the major com~­ age harvestable pines, as well as cause extensive ponents. a-Pinene is thought to be the primary seedling mortality (Davis & Lund 1966; Wilson attractant (Chenier & Philogene 1989). Ethanol is 1968a,b). The pine root collar weevil, H. ,adieU a product of anaerobic plant -and microbial me­ Buchanan, geographically overlaps H. pales and tabolism and is present in varying levels in all pine P. picifJOTtJS, and shares a common host range (Wil­ species. Ethanol levels vary seasonally in Scots pine, son & Millers 1983). However, these larvae breed­ Pinus syltJeStris L., reaching their peaks in late in healthy hosts. Infestation and subsequent tissue April and May (Crawford & Baines 19i7). In ad­ degradation of a healthy host by H. .rad~ may dition to these seasonal changes, ethanol levels rise render trees suitable for subsequent development in response to plant stress (Kimmerer & Koz!owski by H. pales or P. piciV01US. 1982). Traditional control measures consist of preven­ Previous studies of the pine root weevil compie:< tive applications of highly toxic, persistent insec­ have shown that a 1:1 ratio of ethanol to turpentine ticides (Benyus 1983, Drooz 1985), Preplant dip­ in pitfall traps attracts weevils Blank traps or traps_ ping of seedlings or stump applications of lindane with either compound alone show no attraction. are used most commonly. Because of thc sandy Etf-~r:ol :md turpentine act synergistically. and ap­ soils characteristic of much of the pine-producing parently mimic cues used during insect orientation regions of Wisconsin. the threat of groundwater (lUffa & Hunt 1988. Hunt & Raffa 1989). Better contamination from these chemic:lls is gre:lt. Mon­ determination of the cues governing host attraction itoring pine root weevils could potentially reduce could help in developing a monitoring system. insecticide usage to those situations where damage which could le:ld to reduced insecticide use in con- was imminent. -~:olling these weevils. OO-l6-~X;9IiOO-l1!-OO~2S02-00/f) c: 1991 Entomoingic:>i Socletv "f .-\meric:> February 1991 RIESKE'& R.-\FF.-\: PI:\E ROOT \VEE\"lL .\TTRACTIO:\ TO ETH.-\:\OL Tl'RP~:\ . .\lthough ;, 1:1 mi:<ture attracts weevils relative Traps were monitored at ::::::36-h inter.-als in 1958. to ~ntroG. the optimal mi:<ture is unknown. This The monitoring interval in 1989 was appro:<imate­ studv tested the attraction of H. pales and P. pi­ ly weekly, because the volatilization from the larg­ cit;m-us to varying ratios of ethanol and turpentine er container dispensing the 75: 1 ratio .lllowed for in pitfall traps. less frequent replenishment. .\t each monitoring interval. the weevils were removed and the baits were replenished. Weevils were ide~tiaed to spe­ Materials and )'felhods cies and sex using available keys (\Varner 1966. The study was conducted during the summers Wilson et al. 1966, Franklin & Tavlor 1970). of 1988 and 1989, in Waushara County, Wis. A Results from each individual vea~ and both vears study plot was established in the spring of 1988 in combined were analyzed by 'the ge~eral linear a 5\~-yr-old Scots pine Christmas tree farm with model procedure using a split-plot analysis: no. in­ sandy soils. The trees were spaced at 1.68-m in­ sects = block plot block x plot treatment plot x tervili within rows 1.68 m apart. Traps were placed treatment, and Duncan·s multiple range test (SAS in the rows midway between every second tree, Institute 1982). The split-plot analysis was chosen resulting in trap spacing of ==3.36 by ~.36 m. TreeS' to detect any potential plot-treatment interaction. exhibiting foliar ~loration indicative of root A sqUare root transformation was used to normalize weevil infestation were common throughoql, and the data before,. the analysis was performed. Each based on the appearance of the root collar region se:t of each speCies was analyzed separately..-\ XZ of these trees, the infestation level approached 90%. analysis was used to determine if H. pales and P. The pitfall traps were a modification of those picioorus distribution differed across all treatments. used by Hunt & Raffa (1989), and consisted of 17­ A r analysis also was used to assess gender differ­ cm sections of IO-cm diameter plastic PVC drain­ ences across treatments for each species. Each -pipe. Eleven em from one end a series of eight treatment was further analyzed using a test of pro­ 7-mm diameter holds were drilled about the pe­ portions (Snedecor & Cochran 1980). rimeter. The trap interior was coated with liquid Teflon to prevent weevil escape. The traps were Results capped at each end and inserted into the ground so that the holes were flush with ground level Two The seasonal trap catches are shown in Fig. 1. 2-mm holes were drilled in the bottom to allow for Total trap catches wet:e 460 in 1988 and 1,235 in drainage. The exposed 6 cm of the trap was painted 1989. This 2.7 times increase in the number of flat black to simulate a tree trunk image. Baits were weevils trapped included 1.9 times more H. pales dispensed from 2-ml glass vials (0.5 dram. 12 by and 3 times more P. piciV01U3. However, the peak 85 mm) and suspended by thin aluminum wire. A trap catch for both weevil species occurred from stiff 14-gauge wire passed through two 2-mm boles late May to early July in 1988 and 1989. In both in the trap wall, so that the vials were suspended years, P. picioorus showed an additional smaller 4 em below ground level Baits consisted of 95% peak inlate summer that was absent in the H. pales ethaDol and turpentine. The turpentine (Mautz distribution. In 1988, 70% of the weevils trapped _Paint Company, Madison. Wis.) consisted of 46% were P. picioorus, and 30% were H. pales. In 1989, .ci-pmene. -42" p-pinene.,- 2%" p-phelIandreoe. 1% .the distribution was 77 and 23%. respectively. limonene. 0.88" camphene, 0.77" myrcene. and A treatment effect was significant for both sexes <1" unknown compouods,.as·determined by the of each species in 1988 and 1989 (P < 0.001, Table gas-liquid chromotography methodofIWfa & Stef­ 1). There was no significant plot x treatment in­ feck (1988). The volatilization rates under labo- teraction for either species, with the exception of '-ratory conditions:(23"C) were 200 mg/d of ethanol male H. 7'4le3 in 1988 (F = 2.44; d.f = 10, 29; P = and 40 mg/d of turpentine for a 1:1 volumetric 0.01). This single instance of plot x treatment in- ratio. teraction is probably due to the small sampl~ size Bait ratios were manipulated by altering the involved.. There were no other significant inter­ number of vials in each trap containing either eth­ actions in 1988 or 1989. Blocking was not an ef­ anolor turpentine. The ratios were 1:10, 1:5, 1:1, fective means of explaining weevil distribution at 5:1, 10:1, and 75:1. The i5:1 ratio was dispensed this particular site; a significant block effect oc­ from 5-cm-diaioeter Petri dishes in 1988. In 1989, curred only with male P. picivorus in 1989 (F = Petri dishes were rep12ced with 12o-ml glass jars 6.43; d.f "" 4, 8; P < 0.05). (5 by 10 em). The ethanol volatilization rate from Hylobius pale:s were most attracted to the 75:1, --------tnese large! dispel1seu was:--~.l3;OOO mg/d and ..th.~ 10:1, and the 5:1 ratios (Fig.